• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권3호
• /
• pp.1086-1103
• /
• 2020

The automatic identification and classification of image-based weld defects is a difficult task due to the complex texture of the X-ray images of the weld defect. Several depth learning methods for automatically identifying welds were proposed and tested. In this work, four different depth convolutional neural networks were evaluated and compared on the 1631 image set. The concavity, undercut, bar defects, circular defects, unfused defects and incomplete penetration in the weld image 6 different types of defects are classified. Another contribution of this paper is to train a CNN model "RayNet" for the dataset from scratch. In the experiment part, the parameters of convolution operation are compared and analyzed, in which the experimental part performs a comparative analysis of various parameters in the convolution operation, compares the size of the input image, gives the classification results for each defect, and finally shows the partial feature map during feature extraction with the classification accuracy reaching 96.5%, which is 6.6% higher than the classification accuracy of other existing fine-tuned models, and even improves the classification accuracy compared with the traditional image processing methods, and also proves that the model trained from scratch also has a good performance on small-scale data sets. Our proposed method can assist the evaluators in classifying pipeline welding defects.

• 한국BIM학회 논문집
• /
• 제11권3호
• /
• pp.22-33
• /
• 2021

In order to maximize the use of BIM, all data related to individual elements in the model must be correctly assigned, and it is essential to check whether it corresponds to the IFC entity classification. However, as the BIM modeling process is performed by a large number of participants, it is difficult to achieve complete integrity. To solve this problem, studies on semantic integrity verification are being conducted to examine whether elements are correctly classified or IFC mapped in the BIM model by applying an artificial intelligence algorithm to the 2D image of each element. Existing studies had a limitation in that they could not correctly classify some elements even though the geometrical differences in the images were clear. This was found to be due to the fact that the geometrical characteristics were not properly reflected in the learning process because the range of the region to be learned in the image was not clearly defined. In this study, the CRF-RNN-based semantic segmentation was applied to increase the clarity of element region within each image, and then applied to the MVCNN algorithm to improve the classification performance. As a result of applying semantic segmentation in the MVCNN learning process to 889 data composed of a total of 8 BIM element types, the classification accuracy was found to be 0.92, which is improved by 0.06 compared to the conventional MVCNN.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권2호
• /
• pp.751-770
• /
• 2019

Action recognition is an essential task in computer vision due to the variety of prospective applications, such as security surveillance, machine learning, and human-computer interaction. The availability of more video data than ever before and the lofty performance of deep convolutional neural networks also make it essential for action recognition in video. Unfortunately, limited crafted video features and the scarcity of benchmark datasets make it challenging to address the multi-person action recognition task in video data. In this work, we propose a deep convolutional neural network-based Effective Hybrid Learning (EHL) framework for two-person interaction classification in video data. Our approach exploits a pre-trained network model (the VGG16 from the University of Oxford Visual Geometry Group) and extends the Faster R-CNN (region-based convolutional neural network a state-of-the-art detector for image classification). We broaden a semi-supervised learning method combined with an active learning method to improve overall performance. Numerous types of two-person interactions exist in the real world, which makes this a challenging task. In our experiment, we consider a limited number of actions, such as hugging, fighting, linking arms, talking, and kidnapping in two environment such simple and complex. We show that our trained model with an active semi-supervised learning architecture gradually improves the performance. In a simple environment using an Intelligent Technology Laboratory (ITLab) dataset from Inha University, performance increased to 95.6% accuracy, and in a complex environment, performance reached 81% accuracy. Our method reduces data-labeling time, compared to supervised learning methods, for the ITLab dataset. We also conduct extensive experiment on Human Action Recognition benchmarks such as UT-Interaction dataset, HMDB51 dataset and obtain better performance than state-of-the-art approaches.

• 대한방사선기술학회지:방사선기술과학
• /
• 제44권6호
• /
• pp.655-661
• /
• 2021

Positron emission tomography (PET) images is affected by acquisition time, short acquisition times results in low gamma counts leading to degradation of image quality by statistical noise. Noise2Void(N2V) is self supervised denoising model that is convolutional neural network (CNN) based deep learning. The purpose of this study is to evaluate denoising performance of N2V for PET image with a short acquisition time. The phantom was scanned as a list mode for 10 min using Biograph mCT40 of PET/CT (Siemens Healthcare, Erlangen, Germany). We compared PET images using NEMA image-quality phantom for standard acquisition time (10 min), short acquisition time (2min) and simulated PET image (S2 min). To evaluate performance of N2V, the peak signal to noise ratio (PSNR), normalized root mean square error (NRMSE), structural similarity index (SSIM) and radio-activity recovery coefficient (RC) were used. The PSNR, NRMSE and SSIM for 2 min and S2 min PET images compared to 10min PET image were 30.983, 33.936, 9.954, 7.609 and 0.916, 0.934 respectively. The RC for spheres with S2 min PET image also met European Association of Nuclear Medicine Research Ltd. (EARL) FDG PET accreditation program. We confirmed generated S2 min PET image from N2V deep learning showed improvement results compared to 2 min PET image and The PET images on visual analysis were also comparable between 10 min and S2 min PET images. In conclusion, noisy PET image by means of short acquisition time using N2V denoising network model can be improved image quality without underestimation of radioactivity.

• 정보보호학회논문지
• /
• 제32권3호
• /
• pp.501-511
• /
• 2022

2021년에는 코로나의 여파로 랜섬웨어를 활용한 공격이 유행했으며 그 수는 매년 급증하고 있다. 그 중 파워쉘은 랜섬웨어에 주요 기술로 사용되고 있어 파워쉘 기반 악성코드 탐지 기법의 필요성은 증가하고 있으나 기존의 탐지기법은 난독화가 적용된 스크립트를 탐지하지 못하거나 역난독화에 시간이 오래 소요되는 한계가 존재한다. 이에 본 논문에서는 간단하고 빠른 역난독화 처리과정, Word2Vec과 CNN(Convolutional Neural Network)으로 구성되어 스크립트의 의미를 학습하고 특징을 추출해 악성 여부를 판단할 수 있는 딥러닝 기반의 분류 모델을 제안한다. 2021 사이버보안 AI/빅데이터 활용 경진대회의 AI 기반 파워쉘 악성 스크립트 탐지 트랙에서 제공된 1400개의 악성코드와 8600개의 정상 스크립트를 이용하여 제안한 모델을 테스트한 결과 기존보다 5.04배 빠른 역난독화 실행시간, 100%의 역난독화 성공률, 0.01의 FPR(False Positve Rate), 0.965의 TPR(True Positive Rate)로 악성코드를 빠르고 효과적으로 탐지함을 보인다.

• 반도체디스플레이기술학회지
• /
• 제22권1호
• /
• pp.11-17
• /
• 2023

This paper proposes a lightweight deep learning network for detecting an image splicing forgery. The research on image forgery detection using CNN, a deep learning network, and research on detecting and localizing forgery in pixel units are in progress. Among them, CAT-Net, which learns the discrete cosine transform coefficients of images together with images, was released in 2022. The DCT coefficients presented by CAT-Net are combined with the JPEG artifact learning module and the backbone model as pre-learning, and the weights are fixed. The dataset used for pre-training is not included in the public dataset, and the backbone model has a relatively large number of network parameters, which causes overfitting in a small dataset, hindering generalization performance. In this paper, this learning module is designed to learn the characterization depending on the DCT domain in real-time during network training without pre-training. The DCT RRU-Net proposed in this paper is a network that combines RRU-Net which detects forgery by learning only images and JPEG artifact learning module. It is confirmed that the network parameters are less than those of CAT-Net, the detection performance of forgery is better than that of RRU-Net, and the generalization performance for various datasets improves through the network architecture and training method of DCT RRU-Net.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2022년도 춘계학술대회
• /
• pp.275-276
• /
• 2022

해상교통관제사는 항로폭이 협소한 항만에서 선박 충돌사고 예방을 위해 one-way로 항해하도록 선박의 입·출항 대기 지시를 한다. 현재 해상교통관제사의 입·출항대기 지시는 과학적이고 통계적인 데이터를 기반으로 하지 않고 해상교통관제사의 개인 역량에 따라 편차가 크다. 이에 따라 본 연구에서는 항만에서의 선박 입·출항 대기 지시를 위한 정확한 이동 시간을 예측하기 위해 선박 및 기상 데이터를 수집하여 1d-합성곱신경망 모델을 구축하였다. 제안한 모델이 다른 앙상블 기계학습 모델보다 4.5% 이상 개선된 것을 확인하였다. 본 연구를 통해 다양한 상황에서도 선박 입·출항 소요시간 예측이 가능하여 해상교통관제사는 선박에게 정확한 정보 제공 및 대기지시 판단에 도움이 될 것으로 기대된다.

• International journal of advanced smart convergence
• /
• 제12권4호
• /
• pp.434-442
• /
• 2023

The relationship between acute kidney injury (AKI) prediction and nephrotoxic drugs, or drugs that adversely affect kidney function, is one that has yet to be explored in the critical care setting. One contributing factor to this gap in research is the limited investigation of drug modalities in the intensive care unit (ICU) context, due to the challenges of processing prescription data into the corresponding drug representations and a lack in the comprehensive understanding of these drug representations. This study addresses this gap by proposing a novel approach that leverages patient prescription data as a modality to improve existing models for AKI prediction. We base our research on Electronic Health Record (EHR) data, extracting the relevant patient prescription information and converting it into the selected drug representation for our research, the extended-connectivity fingerprint (ECFP). Furthermore, we adopt a unique multimodal approach, developing machine learning models and 1D Convolutional Neural Networks (CNN) applied to clinical drug representations, establishing a procedure which has not been used by any previous studies predicting AKI. The findings showcase a notable improvement in AKI prediction through the integration of drug embeddings and other patient cohort features. By using drug features represented as ECFP molecular fingerprints along with common cohort features such as demographics and lab test values, we achieved a considerable improvement in model performance for the AKI prediction task over the baseline model which does not include the drug representations as features, indicating that our distinct approach enhances existing baseline techniques and highlights the relevance of drug data in predicting AKI in the ICU setting.

• 한국방사선학회논문지
• /
• 제15권5호
• /
• pp.613-620
• /
• 2021

본 논문에서는 흉부 X선 영상에서 정상 심장과 비정상 심장(심장비대)을 분류할 수 있는 합성곱 신경망 모델을 제안하고자 한다. 학습 및 테스트 데이터로는 경북대학교병원에 내원하여 정상과 심장비대를 진단받은 환자들의 흉부 X-선 이미지를 획득하여 사용하였다. 제안된 합성곱 신경망 모델을 이용하였을 때의 정상 심장 및 비정상 심장(심장비대) 분류 정확도는 99.88%였다. 정상 심장 영상을 테스트 데이터로 사용하였을 때의 정확도, 정밀도, 재현율 및 F1 Score는 95%, 100%, 90%, 96%였다. 비정상 심장(심장비대) 영상을 테스트 데이터로 사용하였을 때의 정확도, 정밀도, 재현율 및 F1 Score는 95%, 92%, 100% 및 96%였다. 이러한 학습 및 테스트 분류 결과로 제안된 합성곱 신경망 모델은 흉부 X-선 영상의 특징 추출 및 분류에서 매우 우수한 성능을 보여주고 있다고 판단된다. 본 논문에서 제안하는 합성곱 신경망 모델은 흉부 X-선 영상의 질환 분류에 있어 유용한 결과를 보여줄 것으로 판단되며, 다른 의료 영상에서도 동일한 결과를 나타내는지 알아보기 위하여 추가적인 연구가 이루어져야 할 것이다.

• 한국전자거래학회지
• /
• 제27권1호
• /
• pp.63-79
• /
• 2022

최근 머신러닝 및 딥러닝 기법을 활용한 주식 가격 예측 연구가 다양하게 이루어지고 있다. 그 중에서도 최근에는 주식 매수 및 매도 주문 정보를 담고 있는 호가창을 이용하여 주가를 예측하려는 연구가 시도되고 있다. 하지만 호가창을 활용한 연구는 대부분 가장 최근 일정 기간 동안의 호가창 추이만을 고려하며, 호가창의 중기 추이와 단기 추이를 같이 고려하는 연구는 거의 진행되지 않았다. 이에 본 논문에서는 호가창의 중기와 단기 추이를 모두 고려하여 주가 등락을 보다 정확히 예측하는 딥러닝 기반 예측 모델을 제안한다. 더욱이 본 논문에서 제안하는 모델은 중단기 호가창 정보 외에도 해당 종목에 대한 동기간 뉴스 헤드라인까지 고려하여 기업의 정성적 상황까지 주가 예측에 반영한다. 본 논문에서 제안하는 딥러닝 기반 예측 모델은 호가창 변화의 특징을 합성곱 신경망으로 추출하고 뉴스 헤드라인의 특징을 Word2vec을 이용하여 추출한 뒤, 이들 정보를 결합하여 특정 기업 주식의 다음 날 등락 여부를 예측한다. 실제 NASDAQ 호가창 데이터와 뉴스 헤드라인 데이터를 사용하여 제안 모델로 5개 종목(Amazon, Apple, Facebook, Google, Tesla)의 일일 주가 등락을 예측한 결과, 제안 모델은 기존 모델에 비해 정확도를 최대 17.66%p, 평균 14.47%p 향상시켰다. 또한 해당 모델로 모의 투자를 수행한 결과, 21 영업일 동안 종목에 따라 최소 $492.46, 최대 $2,840.83의 수익을 얻었다.